Directions for Research

It's a familiar parable, but worthy of repeating in this context:

Under a streetlight, on a very dark night, a man was looking for a set of lost keys. A policeman came by and inquired about the object of his search. After the two had searched diligently for quite some time, the officer finally asked, “Are you sure you lost them here?”

“Oh, no,” said the man, pointing out into the darkness. “I lost the keys somewhere over there.”

“Then why are you looking for them over here?” the officer asked.

“Because this is where the light is!” the man replied.

What lies under the streetlight seems largely to have influenced the focus of studies that have examined disparities in breast cancer outcomes. The data collected by cancer registries has offered much that is of value, but much important data still lies largely unexplored in the outer shadows, particularly in the poorly documented areas of incidence and treatment of locally advanced and metastatic breast cancer. Likewise, when it comes to understanding the experiences and needs of advanced breast cancer patients, we have few other informational resources upon which to draw, outside of clinical trials data.

As mentioned earlier in this paper, the incidence and mortality data captured in the California Cancer Registry (CCR)—as in SEER and other cancer registries—offers specific, detailed information at two points in time only, at diagnosis, initial treatment, and death. What happens between these two points in time is largely unexplored territory. Thus, an accurate picture of the extended treatments received by women undergoing aggressive primary treatment, as well as multiple ongoing treatments for metastatic breast cancer are not reflected in this data. Clearly, there are problems with retrieval, accuracy, and analysis of treatment-related data in the CCR and other databases.

This paper has advanced the thesis that the intense focus of the last 20 years on early detection and screening has yielded limited results in reducing absolute rates of stage at diagnosis and mortality. Further, it is clear that this focus has failed to take into account the needs and experiences of women with high-risk and metastatic breast cancer. Few informational resources exist that accurately document the treatments offered to this group of patients. Many believe that it is advances in treatment, as much as (or more than) screening that is responsible for the lowered death rates seen in the last decade or so.

Like the rest of the United States, California has been hampered by poor quality information on breast cancer treatment, a fact that is particularly meaningful for highrisk and metastatic breast cancer patients, whose treatment is by necessity extensive and ongoing. Without accurate treatment data on these vulnerable populations, we will not be able to fully explain differences in outcomes.

Documenting the Patient Population

An important informational gap that exists in cancer registry data is that of the incidence of breast cancer recurrence. Since the database is limited to initial diagnosis, initial treatments and deaths, there is no record of recurrence. This makes it impossible to get a sense of the magnitude of a patient population that includes all patients who ultimately die of breast cancer, except insofar as they can be extrapolated from death rates. Without a clear sense of the numbers and distribution of these patients, and the length of their survival, novel ways to address disparities in the provision of care and other needs are likely to remain elusive.

One subgroup that should be analyzed are those categorized as “unclassified” in the registry data. The outcomes for this group are poor enough that they almost certainly include Stage III and IV patients. It is possible that this group represents a disproportionately underserved and minority population, and that the absence of staging or documentation of staging is an aspect of this lack. An analysis of the reasons why these patients are not staged would be of interest.

Disparities in Outcomes: Issues

The research is confusing when it looks at the issue of race/ethnicity as a predictor of outcomes versus socio-economic status as a predictor. Some studies seem to suggest that race/ethnicity may actually serve as a proxy for SES. A more in-depth understanding of the ways in which these factors play out in California breast cancer populations may provide some important further insight.

Some of the large racial/ethnic subgroupings have very diverse outcomes. Of particular interest are the different races and ethnicities of Asian/Pacific Islanders, whose outcomes vary so widely that larger subgroupings may not be meaningful.

Of particular concern in the Asian/Pacific Islander population, which still overall bears relatively low comparative incidence and mortality rates, are the dramatic recent increases in mortality. While still well under the rates for other groups, the increases have been so rapid that they are worrisome. An investigation of causation, including a comparison of the effects of immigration over time on incidence and mortality, would be crucial.

A number of other issues in understanding disparities in outcome are worthy of further study in the California population. For example, in view of the broad increase in utilization of mammography screening among black women, outstripping that of white women in some subgroups, what role does early detection actually play in explaining the persisting disparities in outcome? If in fact screening and early detection proves to be only marginally effective (or not effective at all) in improving outcomes in certain populations, how well do these current models of intervention actually work? The answers to these important questions will be crucial in allocating scarce funding.

Breast cancer in black women occurs at younger ages, with more aggressive features. Is there a biological difference in the kind of breast cancer that young black women are likely to get? Is there a biological difference in the kinds of breast cancer that certain Asian populations with unusually positive outcomes are likely to get? Are these differences related to inherent factors, to behavioral and environmental differences, or to other factors, as yet undetermined? How do access to treatment, healthcare coverage, patient compliance, and provider biases interact with these differences? These are some of the important research questions that should be examined.

Looking at Treatment

Multiple explanatory factors have been proposed to explain disparities in outcome, including stage at diagnosis, disease characteristics, age, screening behaviors, and treatment variables. Those explanatory factors that are easier to document than others, such as stage at diagnosis and mammography screening, have been more widely studied—the “street light syndrome” mentioned above—while little is known about how, whether, and which treatments are actually offered, accepted or undergone.

Most newly diagnosed breast cancers are hormonally driven, so hormonal drugs like tamoxifen and Arimidex are standard adjuvant treatments. Yet the few studies that have been done suggest that compliance with the full five-year prescribed treatment can be very low in some racial and socio-economic groups. What role do factors like consumer and provider attitudes play in treatment compliance? What role does regular health care, followup appointments, problems with childcare, transportation, and other financial/social issues play in compliance? The issue of compliance with treatment recommendations must be considered in the context of other health care, economic, and social considerations in minority communities.

Patterns of care studies in California women have the potential of following women from all ages, SES, racial, and ethnic groups through diagnosis, treatment, and follow-up to look at whether differences in care received and/or accepted is meaningful. Comparisons between standard of care recommended and actual treatment received should be instructive. Where there is funding for health care in poor and minority populations, for example in women aged 65 and older, has this impacted outcome in a meaningful way in these populations? It would be especially meaningful to examine the number of lines of metastatic breast cancer treatment offered to various groups of patients, as well as the quality of palliative care offered, and actually received by patients.

Are disparities in outcome related to the attitudes, biases, and values of health care providers toward different populations? Could these factors be addressed by physician training in connection with specific areas deemed worthy of remediation, e.g., failure to diagnose breast cancer in young black women, or failure to provide full doses of radiation or chemotherapy to elderly or obese patients?

The issue of treatment costs must be addressed in research, particularly as more patented drugs for the high-risk and metastatic population are approved, costs continue to spiral, and issues of universal access to care are deferred. It would be important to have research that documents the actual dollars spent in treating high-risk and metastatic breast cancer in California, as well as any potential disparities in such expenditures on behalf of different patients across racial/ethnic, socio-economic, and age groups. Understanding how these data may relate to outcome data would serve to clarify some of the fundamental questions raised in this paper. These data would clearly provide important data for health policymakers.

A Final Word…

It's my hope that this report will be of some assistance in bringing into view some of the more challenging needs of California women diagnosed with high-risk and metastatic breast cancer.

As an advocate, I hope one day to see the bright light of evidence shed on even the most difficult social dilemmas—a light so brilliant, with data so clear and persuasive, that we as a people cannot fail to finally address the issues before us, instead of turning away, as we so often do. Clearly, the disparities reflected in the outcome data in breast cancer in California, as elsewhere, point directly to the most difficult social problems we face today: racial and class discrimination, poverty, rising health care costs, diminishing quality in medical care, the continued persistence of life-threatening cancers despite billions of dollars in research, and, of course, the incalculable loss of thousands of women and men each year to breast cancer.

I commend the California Breast Cancer Research Program for your consideration of these challenging issues in breast cancer research. May your program continue to serve as a model, offering support and encouragement to researchers willing to take up this challenge, inspiring them with the dream and the possibility that their research will one day make a difference.